Elac Adante AS-61 Speaker Review

[maty]

About the burning, It will depend on each speaker. I have commented many times that my KEF Q100 (only one coaxial) sounded without bass so I thought about returning them. They needed many hours of burn in (about 200 hours for the sound to stabilize, crazy). My hypothesis is that they used a very rigid coaxial suspension. Presumably, the new versions need less hours but still suffer from this newly released problem.

An example, KEF R5, coaxial + 2 woofers. New the sound is "reasonable" but until bad, at least to me.

 

[maty]

To compare ELAC coaxial with KEF R3 coaxial too:

Compare these two coaxials with other traditional speakers and I hope you notice what the magic of the sound of a good small coaxial means to me. With acoustic instruments and natural voices.

 

[outfaced]

Veery interesting tread ... although i read the forum the last year or so, i register now only because of this tread ... ok, actually the whole speaker review section is promising.
Sooo ... i actually have the AS-61 and had before them LS50, dynaudio special 40 and again KEF ... the R3 that i sold to buy the AS-61.
Sadly measurements don't say much about resolution and even less about joy from listening music (not equipment) and the whole gets even more complicated when we include room acoustics and DSP.
Still it will be very interesting indeed to see measurements from R3.
I am all into coax and 3way in small form factor and still prefer KEFs overall tonality but i wanted something better then the R3. Unfortunately the next there is reference 1 ... a little bit too expensive and too big for my room. Therefor i got the AS-61 - much more refined sound with more details and nuances then R3. Ok bass is weak ... especially regarding the size. Dont see any benefit from this design with the passive radiator. In this volume with classic bassreflex you can get far better bass especially with dedicated bass driver only covering 50-200Hz. In my case not a big deal because of the extra subwoofer but dont understand the whole concept (its pritty difficult to develop something like that).
Still, never got the this 200Hz dip when listening to music. The roloff after 10 Khz is indeed there also in my measurements, but when i fix it with the DSP the heights get too much. So i leave it almost without correction. The dip around 1,5khz is also there and audible, or at least i like it better when i bump the 1-2khz area around +2db.
What i dont see in the measurements but hear is a little bit too harsh sounding trumpets and flute. I think around 6-7khz. Actually is controversial to the measurements i think.
Overall still not the perfect speaker for me, but especially now when discounted (i got it for less than R3 new!) pretty good and better then the R3. Again - its defenitly a matter of personal taste and definitely not for bass-lovers (although i find bass accurate and fast ... just not deep)

 

[maty]

Better than R3 is easy -> a modded R300 or R3. If you opt for the old version (R300), the money saved you invest in the improvement of the inside. Soon they will only be available second-hand.

In any case, it seems to me that even KEF has problems to integrate very well a coaxial and a woofer in the same box (to play very good music recordings).

It seems more logical closed box (or bass-reflex closed) and a couple of good subwoofers cut about 110 - 120 Hz -> like to floorstandings, maybe.

 

[outfaced]

What mode do you have in mind?
The R300/R3 drivers just don't get this fine nuances like the ELAC (or the Dynaudio ... but that is a little bit extreme).
and on the R3 i find the integration of the bass driver actually audible bad. I just hear it like a separate driver.
On the ELAC there is nothing that bothering me regarding integration even the measurement tells another story.

In terms of fast and accurate bass response found the LS50 the best from all, but i am not sure if 70Hz/-3db still counts Now i want to hear them again with a good sub (didn't have one then). R3 goes low, but wouldn't say it's fast or very clean.

 

[napilopez]

Veery interesting tread ... although i read the forum the last year or so, i register now only because of this tread ... ok, actually the whole speaker review section is promising.
Sooo ... i actually have the AS-61 and had before them LS50, dynaudio special 40 and again KEF ... the R3 that i sold to buy the AS-61.
Sadly measurements don't say much about resolution and even less about joy from listening music (not equipment) and the whole gets even more complicated when we include room acoustics and DSP.
Still it will be very interesting indeed to see measurements from R3.
I am all into coax and 3way in small form factor and still prefer KEFs overall tonality but i wanted something better then the R3. Unfortunately the next there is reference 1 ... a little bit too expensive and too big for my room. Therefor i got the AS-61 - much more refined sound with more details and nuances then R3. Ok bass is weak ... especially regarding the size. Dont see any benefit from this design with the passive radiator. In this volume with classic bassreflex you can get far better bass especially with dedicated bass driver only covering 50-200Hz. In my case not a big deal because of the extra subwoofer but dont understand the whole concept (its pritty difficult to develop something like that).
Still, never got the this 200Hz dip when listening to music. The roloff after 10 Khz is indeed there also in my measurements, but when i fix it with the DSP the heights get too much. So i leave it almost without correction. The dip around 1,5khz is also there and audible, or at least i like it better when i bump the 1-2khz area around +2db.
What i dont see in the measurements but hear is a little bit too harsh sounding trumpets and flute. I think around 6-7khz. Actually is controversial to the measurements i think.
Overall still not the perfect speaker for me, but especially now when discounted (i got it for less than R3 new!) pretty good and better then the R3. Again - its defenitly a matter of personal taste and definitely not for bass-lovers (although i find bass accurate and fast ... just not deep)

You're going to have a quite a bit of trouble arguing a speaker more "resolution" around these parts If you hear more detail on the AS-61, barring major distortion, it can almost certainly be explained by frequency response and directivity in some way. I think it's likely the AS-61 simply highlights some sounds more than others on some tracks and that you might interpret that as showing more 'nuances.'

My measurements of the R3:

The two speakers I've overlapped with Amir's measurements have been pretty close, so I think I can roughly compare broader tonal qualities.

Now, the R3 are very good, especially 15 degrees off axis or so. Also note my scaling exacerbates issues relative to amir's since I use a taller aspect ratio. Nonetheless, the R3 are lacking some midrange energy relative to the andante and that will lead the Elacs to sounding more resolving on some tracks.

Edit: To be clear, there's nothing wrong with you preferring the andante. Just what we perceive as "resolution" can usually be tied back to frequency response and directivity in some way. We are sensitive to quite small changes.

 

[Spocko]

In this channel you can listen and compare with other speakers, specially others coaxials like KEF or the substitute, ELAC Carina.

https://www.youtube.com/channel/UCLleWZ74MYiSxG31BkmBnvQ/videos

I've seen this channel, but I always wondered whether the output is properly level matched? Even a 1 dB difference would give the louder speaker a significant advantage in sound quality. If the answer is NO, then the comparisons are practically useless as that means any noticeable differences are likely loudness differences rather than sound quality.

 

[Spocko]

To compare ELAC coaxial with KEF R3 coaxial too:

Compare these two coaxials with other traditional speakers and I hope you notice what the magic of the sound of a good small coaxial means to me. With acoustic instruments and natural voices.

What do you think of the Genelec Ones coaxial - supposedly state of the art in terms of horizontal dispersion and ruler flat frequency response that can be tailored by GLM to your exact preferences.

 

[gr-e]

My measurements of the R3:
View attachment 50723

Notice the lack of diffraction from the edge of midrange driver that the AS-61 and IN-8 have

 

[Soniclife]

The thing is, 200Hz is an absolutely typical frequency at which to cross over a mid in a 3-way.

In my casual observation 350Hz upwards seems most common, I'm not sure I seen anything normal cross over at 200Hz. I checked a random Kef, they used 400, if this crossed over at 400 the midrange looks like it would in it's comfort zone.

 

[napilopez]

Notice the lack of diffraction from the edge of midrange driver that the AS-61 and IN-8 have

Yeah, you can see the advantage of the KEF coaxial. It acts as a much smoother waveguide. Slightly off axis is still better but still much less of the on-axis messiness normally seen in coaxials.

 

[andreasmaaan]

In my casual observation 350Hz upwards seems most common, I'm not sure I seen anything normal cross over at 200Hz. I checked a random Kef, they used 400, if this crossed over at 400 the midrange looks like it would in it's comfort zone.

Different companies/designers have different tendencies. Revel for example tend to cross their 3-ways over at 200-250Hz-ish. Wilson is another one that comes to mind (not to suggest anyone should follow their lead on any design matter, ha). I do agree that 300Hz+ may generally be a better crossover frequency in many cases.

 

[infinitesymphony]

I borrowed these speakers to test them out and I was also not a fan. I think the internal driver to passive radiator didn’t work out nearly as well in reality as it did in theory. Bass seemed bloated, slow and not nearly as impactful as it’s size would imply. It also didn’t seem to time align with the tweeter and mid very well. All in all, I was not a fan.

This touches on a question I've had since Ethan Winer heavily repped the supposedly flat frequency response of studio staple Mackie HR824. I've used these monitors a number of times--a friend has a pair--and while the response seemed reasonably flat, I found the bass hard to figure out. I would use similar adjectives: slow, tubby, indistinct. It uses a passive radiator in the rear pointed outward from what I recall.

Is time-frequency delay something the Klippel measurements would reveal? Have to imagine it would be, waterfall or otherwise. Apologies if this has already been discussed elsewhere.

 

[andreasmaaan]

This touches on a question I've had since Ethan Winer heavily repped the supposedly flat frequency response of studio staple Mackie HR824. I've used these monitors a number of times--a friend has a pair--and while the response seemed reasonably flat, I found the bass hard to figure out. I would use similar adjectives: slow, tubby, indistinct. It uses a passive radiator in the rear pointed outward from what I recall.

Is time-frequency delay something the Klippel measurements would reveal? Have to imagine it would be, waterfall or otherwise. Apologies if this has already been discussed elsewhere.

Do you have a link to any measurements of the HR824? What you're hearing could be the result of any number of things, speaker- and/or room-related.

In any case, it can be inferred from the design that the woofers roll off acoustically at approximately 36dB/octave, and in addition to that there's an electrical high-pass protection filter adding an additional12dB/octave roll-off below the tuning frequency (or more, the setting is user selectable).

So, in the best case scenario (approx. 36+12=48 dB/octave @37Hz) group delay is likely to be in the range of 20-30ms (or slightly more) at 37Hz, and probably with quite a high Q too. There hasn't been adequate research done IMO to determine whether such a degree of group delay might be audible. But it's conceivable.

OTOH, there are various other variables - in particular, the room - that are probably stronger candidates for the sound you describe.

 

[infinitesymphony]

Do you have a link to any measurements of the HR824? What you're hearing could be the result of any number of things, speaker- and/or room-related.

None that I would trust. It's such a classic studio workhorse speaker that I hope it gets tested along with the other favorites from the last 30 years (Genelec, Dynaudio, Yamaha, Event, et al). There have also been debates about the two revisions: made in USA for a number of years, made in China for the rest.

OTOH, there are various other variables - in particular, the room - that are probably stronger candidates for the sound you describe.

Most definitely, I have never heard these speakers in a treated room and I imagine the passive radiator's distance from the wall could have major consequences as well.

 

[infinitesymphony]

So, in the best case scenario (approx. 36+12=48 dB/octave @37Hz) group delay is likely to be in the range of 20-30ms (or slightly more) at 37Hz, and probably with quite a high Q too. There hasn't been adequate research done IMO to determine whether such a degree of group delay might be audible. But it's conceivable.

Thinking about this a bit more...

20-30 ms is quite a lot. In the visual world, that would be almost 1 frame at 30 fps. A display with 30 ms input lag would be considered unusable by competitive gamers. Light travels faster than sound and uses a separate processing system so I understand that this analogy is not 1:1.

In the audio world, it looks like the average person can hear delays 15 ms and greater, with musicians able to notice 5-10 ms delay (hence the importance of low-latency monitoring in audio interfaces).

At 30 ms delay, sound from the drivers could travel up to 33.75' before the radiator begins to vibrate, right?

(1,125 ft./sec / 1000 ms/sec * 30 ms = 33.75')

 

[edechamps]

In the audio world, it looks like the average person can hear delays 15 ms and greater, with musicians able to notice 5-10 ms delay (hence the importance of low-latency monitoring in audio interfaces).

I very much doubt that's true at 37 Hz, which is the frequency @andreasmaaan is referring to.

 

[andreasmaaan]

Thinking about this a bit more...

20-30 ms is quite a lot. In the visual world, that would be almost 1 frame at 30 fps. A display with 30 ms input lag would be considered unusable by competitive gamers. Light travels faster than sound and uses a separate processing system so I understand that this analogy is not 1:1.

In the audio world, it looks like the average person can hear delays 15 ms and greater, with musicians able to notice 5-10 ms delay (hence the importance of low-latency monitoring in audio interfaces).

At 30 ms delay, sound from the drivers could travel up to 33.75' before the radiator begins to vibrate, right?

(1,125 ft./sec / 1000 ms/sec * 30 ms = 33.75')

As @edechamps said IMO.

FWIW, the extent of a delay that is audible is dependent on multiple factors. We're actually extremely sensitive to interaural delays, that is, when a sound arrives at one ear later than it arrives at the other - and especially in low-midrange frequencies. This is related to our hearing system's use of interaural delays to determine the location of sounds in the horizontal plane. Our level of sensitivity to these kinds of delays can be orders of magnitude below 1ms, depending on the frequency and envelope of the sound.

Following from this, we are very sensitive to interchannel delays in stereo reproduction. Delay one channel by a relatively tiny degree, and the image will shift towards the other channel (this can be compensated partially, but not fully, by increasing the level of the delayed channel btw, since we also use interaural SPL differences to localise sounds - although our sensitivity for interaural level differences is greater in the higher frequencies).

However, we are far less sensitive to delays that effect only some frequencies (of both or all channels) equally. This type of delay is called non-constant group delay, and is the type of delay we have been talking about here. As I mentioned before, this area has not been thoroughly researched with respect to low bass frequencies (to my knowledge). However, I can give you some general bits and pieces of info from the research that I've read:

• The audibility of non-constant group delay is highly frequency dependent. We appear to be most sensitive to it in the (lower) midrange.
• The amount of delay in ms is not alone a reliable predictor of its audibility. Also important is the Q of the phase shift, or in other words, how sharply the delay occurs between frequencies. To illustrate this: If non-constant group delay occurs gradually from top to bottom of the entire audio spectrum, it will be very hard to detect, even if the total difference between the arrival time of lowest and highest frequency sounds is tens of milliseconds. By contrast, a lower-magnitude delay that occurs within a very narrow frequency band is much more likely to be audible. This is because we are more sensitive to delays within a particular "critical band" (see explanation here), and highly insensitive to delays between critical bands.
• We are more sensitive to non-constant group delay when listening on headphones (or presumably with speakers in anechoic chambers) than with speakers in normal rooms with reflections.
• In testing, subjects have shown sensitivities to delays of as little as 1 to 2ms in the frequency range 200-1000Hz using headphones, but appeared to be far less sensitive to much larger group delays at lower and higher frequencies.

I've ABX-tested myself for group delay sensitivy using headphones and music recordings and have not been able to detect delays corresponding to a phase shift of one cycle (equivalent to a 24dB/octave filter) at any frequency, or at multiple frequencies in different parts of the audio spectrum in the same stimulus. For example, I was unable to detect the group delay caused by a phase shift of 3 cycles (each cycle corresponds roughly to a 24dB/octave filter) placed in three separated bands in the audio spectrum (from recollection, each "filter" was separated by around two octaves, with the first at about 40Hz). I haven't tested myself for more dramatic delays than that since these don't normally occur in audio reproduction, except occasionally in the low bass when a ported or passive radiator system is electronically high-pass filtered, and even then the magnitude of the delay is not more than two cycles (as is probable in the case you described), except in highly-contrived or unusual circumstances.

I can probably track down some of these other studies I mentioned if you're interested.

 

[WonderBoy]

I hope Andrew chimes in on these... he may have been going for something -- like he's trying to fake his TAD ref1's (fast beryllium drivers, both the concentric AND bass, the light 8-inch passive) and ran out of time. These are really good/interesting/impressive if you work with them... bass is better moving them close to the rear wall (they are sealed, no port). Maybe it's just freq. response, but there's low mid bass detail/openness I never hear anywhere else. Also I swear they are breaking in, off-putting at first, then they loosen the hell up and sweeten. Interesting to say the least.

There may have been a revision... latest SN's (I think) have a single internal port vs. two, but could be minor.

 

[amirm]

I hope Andrew chimes in on these...

He responded and said he would....